as compared with cotton or wool must be owing much less to its own greater specific conducting power than to the smoothness and inelasticity of its fibers. It has too slight a hold on the included air. The more finely shredded it is the better it works; but our experiments have proved that it is not to be recommended as a non-conductor. And yet asbestus is often spoken of as though its excellence in this respect were unquestionable; but, because this wonderful mineral is very useful in many ways by reason of its incombustibility, it does not follow that it has any magic virtue in its other relations to heat. Asbestus paper intercepts heat somewhat better than the loose fiber; but a great many layers must be put together, and then the virtue is by no means commensurate with the cost. It is sometimes recommended as a suitable article to put between floors to prevent the spreading of a possible fire; but those who propose it for this use seem to overlook the fact that the efficiency of non-conductors is nearly proportional to their thickness, and, though an inch might be of some service, one fiftieth of an inch can do very little good.
Fibrous matters and powders in the loose state are somewhat troublesome to confine in the form of coverings, and hence they are sometimes consolidated into sheets or blocks which can be handled without breaking and applied easily. Hair-felt, which is made in thick sheets from the hair which tanners scrape from hides, is cheap and is very serviceable when the heat is not scorching. Paper pulp has been formed into very thick, hollow, half cylinders to put around steam-pipes. Carbonate of magnesium and fossil meal cohere when moistened and slightly compressed, and they may be made into slabs with the addition of a very small percentage of hair or asbestus to give toughness. Such a paste may be plastered directly on steam pipes or boilers and allowed to dry, the fiber serving to prevent cracking; but the greater compactness of light materials so consolidated renders them less effective, especially when a heavy cementing substance is added, like clay or plaster of Paris.
Of non-conducting substances that are already in the solid form, the light woods are often used advantageously. It should be noticed that most of them conduct heat much better along the grain than across it. Thus a cross-section of Liriodendron, or yellow poplar, was found to transmit heat nearly twice as fast as a board of the same thickness sawed lengthwise. Cork is expensive and hard to get in large pieces; but it is far preferable to wood, as it is lighter and more elastic and does not absorb water. Very porous and light bricks confine heat much better than those that are hard burned, but they must be kept dry.
The presence of moisture in a non-conductor greatly impairs its usefulness, as every one knows who has attempted to hold a
